A Selective Multiclass Support Vector Machine Ensemble Classifier for Engineering Surface Classification Using High Definition Metrology

Abstract

The surface appearance is sensitive to change in the manufacturing process and is one of the most important product quality characteristics. The classification of workpiece surface patterns is critical for quality control, because it can provide feedback on the manufacturing process. In this study, a novel classification approach for engineering surfaces is proposed by combining dualtree complex wavelet transform (DTCWT) and selective ensemble classifiers called modified matching pursuit optimization with multiclass support vector machines ensemble (MPOSVME), which adopts support vector machine (SVM) as basic classifiers. The dualtree wavelet transform is used to decompose threedimensional (3D) workpiece surfaces, and the features of workpiece surface are extracted from wavelet subbands of each level. Then MPOSVME is developed to classify different workpiece surfaces based on the extracted features and the performance of the proposed approach is evaluated by computing its classification accuracy. The performance of MPOSVME is validated in case study, and the results demonstrate that MPOSVME can increase the classification accuracy with only a handful of selected classifiers.